getting MT3 tools to release from the taper on mill spindle

[Michael GilliganParticipant @michaelgilligan61133]

Some great photos from Timken, here: **LINK**

https://www.timken.com/pdf/5892_Bearing%20Damage%20Analysis%20Brochure.pdf

Brinelling [see p19] is what we risk when forcibly releasing our stuck Morse Tapers.

MichaelG.

[andrew lynerParticipant @andrewlyner71257]

Whenever I hit the end of a threaded component, in order to extract it, I am worried about the longitudinal stress put against the thread – particularly when the thread may have been partially undone. No one in the thread seems to have considered this. Am I worrying too much? I am not King Kong but I have seen visible signs of a thread being distorted and also evidence that the feel of the thread changes as you go on and off the distorted section.

I appreciate that a draw bar is not a particularly vital part and easily replaceable but it does seem to be the poor relative in tool attachment.

[not done it yetParticipant @notdoneityet]

I would think those engineers of 80 years ago knew that thumping with a hard steel, hammer damaged bearings and a soft babbit mallet did not. Clearly not good enough evidence for some of our modern-day experts on the forum, who seem to think differently. Little wonder that modern day machines need early bearing changes and ‘up-grades’ if people are damaging the originals in this manner – then likely blaming chinese bearing quality.

I’ll carry on using a softer device than a hard steel hammer. The rest of you can make your own choice. As I said earlier I use wedges most of the time if more than a tap with a soft mallet is required to dislodge a taper.

As for the matter of wedges through drilling machines, the angle of the wedge is made sufficiently shallow such that the force is mostly vectored in the direction of the tooling, rather than on bearings. Very basic inclined plane technology often used by civilisations of long past.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by not done it yet on 22/04/2020 23:43:44:

[…]

As for the matter of wedges through drilling machines, the angle of the wedge is made sufficiently shallow such that the force is mostly vectored in the direction of the tooling, rather than on bearings. Very basic inclined plane technology often used by civilisations of long past.

.

Only true if the wedge is actually moving

… A point that I addressed in my reply to Peter G Shaw

MichaelG.

[HopperParticipant @hopper]

I'm still not convinced, regardless of what the 1940s tractor mechanics of Wisconsin thought.

Modern bearing manufacturers all are quite adamant that no pressure, from even a press or puller let alone a hammer, should be put through both races and the rolling elements of a bearing during installation etc. as it can damage balls/rollers and races. Seems to me that a self-eject drawbar that attaches to the spindle itself (not the belt cover or machine body) and does not put pressure through the bearing assembly would be the best way to go. Hammers, soft or hard, are doing exactly what bearing manufacturers specifically say not to do to their product.

The Wisconsin tractor engine flywheel removal pics are somewhat quaint by modern standards. Why would they not put two threaded holes the flywheel so a puller could be attached with a screw bearing on the end of the shaft to pull the flywheel off? Or even cast in a ridge for a standard claw puller to fit on to?

Edited By Hopper on 23/04/2020 02:28:08

[Michael GilliganParticipant @michaelgilligan61133]

Posted by not done it yet on 22/04/2020 23:43:44:

I would think those engineers of 80 years ago knew that thumping with a hard steel, hammer damaged bearings and a soft babbit mallet did not.

.

Please allow me a variation upon that theme :

I would think those engineers of 80 years ago knew that thumping with a hard steel, hammer damaged bearings and presumed that thumping with a soft babbit mallet did not.

MichaelG.

Edited By Michael Gilligan on 23/04/2020 07:21:28

[JasonBModerator @jasonb]

Or was it just that the "soft" hammer did not damage the actual item being hit eg shaft, drift or drawbar like a hardened steel hammer face would and the actual act of hitting was not a lot different between the two.

[Ex contributorParticipant @mgnbuk]

Self ejecting draw bars must also exert a force on the bearing

No. The whole idea of an ejecting drawbar is that it puts not load on the spindle bearings.

On the Emco FB2 & clones. the drawbar has a collar – a cap is screwed on the the end of the spindle that retains the drawbar within the spindle, free to rotate but axially constrained . To release a tool, the spindle is held with a spanner on flats on the spindle nose & the drawbar unscrewed – when the collar on the drawbar touches the cap, further untightening generates an ejecting force within the spindle that does not affect the spindle bearings in any way. As the retaining cap has a finer pitch thread than the drawbar, the cap does not unscrew.

Hitting bearings (or shafts held in bearings) can never be considered a good idea.

Nigel B.

[Clive FosterParticipant @clivefoster55965]

Fundamentally grip in the taper will be down to how tightly the drawbar is done up.

25 to 30 ft lb is said to be ample for R8, both collets and tooling, with no risk of cutter slippage in normal use. I believe Bridgeport call for 28 ft lb and Tormach 30 ft lb. I use about 25 ft llb.

However R8 is a non-holding taper and inherently has less friction in the spindle than the self holding MT3. So an R8 "ought" to need more torque to hold a tooling arbor but less to grip a tool or workpiece in a collet.

Objectively MT collets are a flawed design as the high friction inherent to the self holding geometry reduces the grip on the tool or work for any given drawbar torque. To make matters worse the higher drawbar torque pulls the collet even harder into the taper making extraction disproportionately more difficult. Couldn't have a system better arranged to promote over-tightening. But less than ideal engineering frequently has to be accepted when striking a balance between affordable and adequately functional.

Doesn't help that MT spindles tend to be found on smaller, lighter machines inherently less able to cope with abuse.

If I had an MT taper vertical mill I'd standardise on Weldon type side-lock holders. There aren't that many sizes of cutter shank to deal with. The side-lock holder is shorter than the equally, possibly even more, effective but rapidly becoming obsolete Clarkson screwed shank system. Probably little more than finger tight on the drawbar would be fine, maybe 6 to 10 ft lb. Its a pity more attention hasn't been paid to making really short side-lock holders for small machines. Grinding a flat on the cutter is a pain tho'.

Given that the number of times tooling has been removed from taper spindles by hitting the drawbar over the years without apparent ill effect must be well beyond convenient, or even inconvenient, counting its clear that the procedure is, in practice, acceptable. The engineering and theoretical objections are perfectly valid but if correctly done and the drawbar not over-torqued "not best practice" doesn't reach the level of abuse. As is so often the case the rules require interpretation for special circumstances. But you'd be well advised to stick to them if not prepared to properly investigate the consequences of going outside them.

Clive

[andrew lynerParticipant @andrewlyner71257]

Where does all this lead us? MT is not ideal but I wonder how it came to be so popular. Despite what experienced members have to say about it, the flaw is not enough to blight the whole stable of mini lathes with dodgy bearings that we should, according to local experts, all be suffering. It it really were a major issue then why is there not a source of suitable pullers advertised and available from a host of Eastern entrepreneurs? The puller is clearly a perfect solution if there really is a problem – perhaps better than the captive drawbar which is not an easily retro-fit.

There must be some strong feelings about, judging by the length of this thread. (Keeps a Forum lively).

I have a similar problem with my retro style Pavoni Europiccolo lever espresso coffee maker. That has a long lever which is 'balanced' by the column of the boiler. A design that's looked upon as weak and scary (a split in the boiler could spray you with boiling water) but which is still used and unmodified for decades. A strengthening mod would be easy to incorporate but it would spoil the whacky appearance, perhaps.

[Martin KyteParticipant @martinkyte99762]

Posted by mgnbuk on 23/04/2020 07:47:55:

Self ejecting draw bars must also exert a force on the bearing

No. The whole idea of an ejecting drawbar is that it puts not load on the spindle bearings.

Nigel B.

Think about it.

Whilst the taper is still tight the spindle the toolholder/drawbar assembly are essentially one piece locked together by the taper.The spindle is constrained from moving by the lower spindle bearing. Force is applied to the top of the drawbar putting a downwards force on the spindle and taper tooling assembly.As the spindle is constrained the bit that gives is the taper. If no opposing force is applied to the outer part of the spindle, in our case through the lower spindle bearing the taper CANNOT part company with the spindle. Try ejecting a taper with the quill unlocked and all that happens is the spindle is moved downwards. Therefor self ejecting drawbars must also exert a force on the lower spindle bearing albeit in a controlled way.

regards Martin

[Michael GilliganParticipant @michaelgilligan61133]

Posted by andrew lyner on 23/04/2020 12:06:49:

Where does all this lead us? MT is not ideal but I wonder how it came to be so popular. […]

.

It’s a ‘standard’ that is cheap, convenient, and compact

[ Which is nice for those implementing it ]

• Good for some applications
• Questionable for others

But what do they care ?

It can be marketed on the merit of ‘one system for all the machines in your workshop’.

MichaelG.

[Martin KyteParticipant @martinkyte99762]

If however the design of the ejector is such that it exerts a force betwwen the spindle and the drawbar then yes the bearings are not affected. The retrofit designs I've seen do not appear to do this.

regards Martin

[Ex contributorParticipant @mgnbuk]

Try ejecting a taper with the quill unlocked

No problem on an FB2.

If however the design of the ejector is such that it exerts a force betwwen the spindle and the drawbar then yes the bearings are not affected.

Which was what I thought I had described ? Must try harder !

MT or R8 ? Can't say I have a preference. I use a an XYZ Bridgeport clone at work & it works fine. I use a 2MT FB2 clone at home & it works fine. I have a reasonable selection of 2MT tooling, so wouldn't look to change to R8 particularly if I wanted to change the machine as I would have the added expense of re-tooling.

MT spindles are possibly more compact than R8 for a small machine – the recent thread on spindle runout issues suggests that one style of machine in R8 flavour may have a weak spindle as it was originally designed as MT & the change to R8 has left insufficent material.

Nigel B.

[Clive FosterParticipant @clivefoster55965]

Like Andrew I've often wondered how MT collet systems came to be popular.

MT aren't the only collet system based on self holding taper angles. For example the Brown & Sharpe system that used to be popular, especially in America, has angles close to Morse and is certainly self holding. Brand name B&S collets are invariably partially cut away so that most of the spindle bore contact is in the area needing to be closed down to hold a cutter with a smaller region at the top to ensure concentricity. Typically only around half the collet length actually contacts the spindle. Which must significantly reduce the self holding forces trying to keep the collet in the spindle and increase the percentage of drawbar force acting to hold the collet.

Of course the B&S collets first came out when spindles were mostly plain bearing so the impact issues were clearly of lesser consequence.

Where roller or ball bearings are concerned the general blanket condemnation of applying impact loading through the balls or rollers is primarily to avoid confusion. Far better to simply ban a normally unnecessary process than to confuse the less initiated user with a laundry list of when, and how much, might be safe. When it comes to spindle bearings you can rest assured that any competently engineered system can cope with any impact loads applied in proper operation. Which includes whacking the drawbar to remove a properly inserted collet or tooling arbor. Gross over tightening or otherwise jammed solid for some other reason not counting as proper insertion and operation.

Clive

[egaParticipant @ega]

I used to tap the drawbar on my mill/drill until I became aware of an easy way to make a self-extracting arrangement. Here is the quill complete with standard drawbar:

And here is the new drawbar modified by the addition of a fine thread under the hex head:

The spindle within the quill is, of course, tapped to suit.

In use the taper tool is screwed on to the lower drawbar thread and when it contacts the female taper it is nipped up by hand via the drawbar hex. To release, the hex is undone by about a quarter turn and the differential pitch between the two threads causes the taper to release.

So far as I can tell, the bearings are not involved in this action.

Tooling with a tang rather than a thread is extracted via the slot in the quill.

[Peter G. ShawParticipant @peterg-shaw75338]

In the George McClatchie version for the Sieg X2 mill, the top of the spindle happens to be threaded. George's idea is to create a cap which screws onto this threaded portion. Through the top of the cap is a threaded hole through which a bolt is screwed to bear onto the top of the drawbar which in turn presses the MT tool downwards. Thus the downwards force to release the taper is wholly within the spindle. Rotational movement of the spindle is prevented by means of a pin which is inserted through a hole in the supporting frame into a hole in the spindle itself thus the rotational force of freeing the MT tool is taken directly by the frame.

I suppose in this version there will be some force exerted onto the top bearing, but I cannot see it amounting to much.

In the version I designed for my lathe, the drawbar has a flange which exerts pressure on the tail end of the spindle (mandrel) when tightening the MT tool, in my case a collet, and thus draws it in. To release the MT collet, I screw a cap, similar to McClatchie's idea, onto the exposed thread at the tailend of the spindle (mandrel). This cap, rather than having a bolt through the end, actually bears on the other side of the drawbar flange and thus forces it inwards to exerts pressure directly onto the MT taper via the drawbar. In this instance all forces are constrained totally within the spindle (mandrel).

It must be pointed out that I have found that I can release the taper by holding, eg, the chuck, with the right hand, whilst turning the extractor cap at the other end by using the left hand. I suppose there will be some minimal force onto the bearings whilst doing this but considering it is handheld, I cannot see it amounting to much.

In the McClaskie version, the top spindle securing nut has to be slimmed to enable there being enough thread for the releasing cap to be screwed on. In the case of my lathe, there is approximately 10 to12mm of exposed thread on the end of the spindle (mandrel) which means that no modifications are made to the lathe.

Peter G. Shaw

[SillyOldDufferModerator @sillyoldduffer]

Posted by Michael Gilligan on 23/04/2020 12:24:41:

Posted by andrew lyner on 23/04/2020 12:06:49:

Where does all this lead us? MT is not ideal but I wonder how it came to be so popular. […]

.

It’s a ‘standard’ that is cheap, convenient, and compact

[ Which is nice for those implementing it ]

• Good for some applications
• Questionable for others

But what do they care ?

It can be marketed on the merit of ‘one system for all the machines in your workshop’.

MichaelG.

It's the first quick-change system invented and it's stood the test of time! As I said earlier, there isn't much evidence that releasing MT with a hammer is a problem. As a system it works provide the tapers are kept clean and the drawbar correctly tensioned. I agree with Michael that MT isn't a universal panacea, but neither are the alternatives. Most systems are deranged by dirt and abuse.

I suggest MT's main disadvantage is the extra time taken to swap tools compared with other quick-change systems. Important where speed matters but it doesn't rock my world much. The Morse taper's friction grip is appropriate to the forces applied by machines fitted with it. R8 was invented to be faster and to transmit more power, but 'more power' in that context means Bridgeport and up. Is extra power handling really important on a small mill?

I bought MT throughout to enable tool exchange between my mill and lathe. It's only a mild convenience, but standardising is good reason for choosing MT, or any other system. It wouldn't have been smart to buy an MT lathe and an R8 mill unless there was advantage in doing so.

One good reason for going R8 is a need for extra power handling; I don't have that requirement – my mill is only 1100W. Another sensible reason would be access to lots of cheap R8 tooling. Maybe if I was a machinist working in the USA. As it is, for me in the UK, R8 doesn't have a cost or availability advantage over MT.

Not sure other people saying "R8 is better than MT" is any recommendation at all. Hearsay is often duff, best taken with a large pinch of salt or not at all. Generalisations can be deeply misleading unless the underlying reasons happen to match your particular circumstances rather than theirs. Exactly why is R8 better for you? Only go for it if there's a good answer!

Dave

[Anonymous]

As far as I'm aware the Morse taper was mostly used for drills, where the loads are predominantly axial. Some old school milling machines used it, but not many of the professional mills; they preferred the INT series. Some smaller manufacturers used special standards such as Brown & Sharpe or Jarno. But I suspect that was more to do with forcing the customer to buy your collets rather than any technical advantage. In the model engineering world I suspect it was lazyness rather than any technical reason to use Morse tapers on milling machines; plus repeating the mantra that you can use the same tooling as the lathe. Personally I don't see that. The only common tooling I'd have is drill chucks, and I'd rather have chucks for each machine than faff about sharing them. Once in a blue moon I use Morse taper drills or reamers on the vertical mill, for which I have an adapter. For anything other than drills you need a different style anyway, tang versus drawbar. So sharing tooling is even less useful.

My vertical and CNC mills have R8 tapers, although for most work on both I use the Tormach quick change system. This uses tool holders with a 3/4" shank which fit into a modified 3/4" collet, which normally stays in place. On the CNC mill I can't remember the last time I removed it. It gets removed more often on the vertical mill, last time was to trial the ARC insert milling cutters.

Given a choice I'd use the INT40 taper I have on the horizontal mill. It's rigid and can transmit real power while being self releasing. Power is transmitted via dogs and slots in the arbor rather than friction. It knocks both Morse taper and R8 into a cocked hat.

Andrew

[egaParticipant @ega]

Posted by Andrew Johnston on 23/04/2020 17:07:19:

It knocks both Morse taper and R8 into a cocked hat.

Andrew

I first came across that expression in Philip Gibbs' Darkened Rooms published in 1929; that hat must be quite full by now!

[Howard LewisParticipant @howardlewis46836]

My "extractor" for the Mill is, in essence, the same as that described by peter Shaw, in tyhat all forces are contained within the quill, exerting no force on the bearings.

Rereading the OP, it refers to a mini lathe.

In this case, I had already made an extension to the mandrel, to carry any swarf outside the gear cover. This was drilled and tapped in three places for M3 (largest I felt could be fitted in ) grubscrews.

The clamping nut fore the drawbar was undercut, so that when the three grubscrews were fitted and protruding to the undercut, slackening the drawbar exerted a force pushing the drawbar and 3MT fitting out of the Mandrel. Again, without placing any forces on the bearings.

It would be better if the undercut were a little wider, and the grubscrews a little larger to match, to ensure adequate strength.

Just my way of solving that particular problem.

A more elegant way would be to screw the extension onto the mandrel and make things larger, buthaving jumped, I had to swim as best I could.

Howard

Edited By Howard Lewis on 23/04/2020 17:48:43

[egaParticipant @ega]

Posted by Howard Lewis on 23/04/2020 17:47:18:

Rereading the OP, it refers to a mini lathe.

Good point, which I had overlooked in my earlier post . I note also that andrew lyner seems to have solved his problem with a slide hammer operating directly on the tail of the taper tool.

Lathes with threaded spindle noses could exploit the thread to carry a flanged cap which when unscrewed would engage with the taper tool and thus extract it – don't Myford collets use this feature?

[Martin KyteParticipant @martinkyte99762]

Well I partially retract my comments, the designs I've seen, not that I looked at many, used the spindle cover which is attached to the mill head not the spindle. I would still like to hear any comments that the force on the bearings is somewhat self limiting when giving the drawbar a tap.

regards Martin

[Martin ConnellyParticipant @martinconnelly55370]

This is not a problem I have as the only morse taper tooling I have is mounted in a tailstock which ejects on retraction. I was wondering though if adding a large mass to a lathe spindle in the form of a chuck would help if hammering a draw bar? On a mill could a bridge be set up to support the bottom of the spindle against a heavy plate on the bed before hitting with a hammer?

Martin C

[andrew lynerParticipant @andrewlyner71257]

Posted by Martin Kyte on 23/04/2020 12:21:25:

Think about it.

Therefor self ejecting drawbars must also exert a force on the lower spindle bearing albeit in a controlled way.

regards Martin

That doesn't follow at all because it all depends on where both forces are applied. It is logically possible (not always practical though) to make sure that the two opposing forces are only applied to the taper and spindle. True, It would be possible to push against the top of the drawbar but, when there is room in 'that region' at the top of the spindle for extra threads and a sleeve of some sort, the forces could be balanced.

On another matter, people seem to be claiming that dynamic and static forces on a bearing would have equivalent effects. (i.e. bashing = pressing) That can't be true in general. Hitting one end of a linkage can produce very different forces than the maximum force from the hammer. Overcoming static friction is a matter of peak force whereas the amount of damage done is more related to the Work Done. The follow through from a heavy hammer will involve all the original Kinetic Energy of the hammer, long after the taper has been dislodged and the KE could end up being transferred into distortion (working the metal). A heavy hammer must be restrained from hitting the spindle after the taper has left it. Can we be sure of achieving that? A rigid puller exerts virtually no force, once the taper has moved by less than 1mm.

Impulse is defined as Force times time, where Work is Force times distance. Trying to reconcile these two things is a problem that you will find in sport, insurance claims and manufacturing Science. Finding the appropriate approach will depend on the specific application and involves suppressing intuition. Most of the race memory about these things in hobby Engineering is based on situations involving 'old materials'. Modern bearings are not made of old materials so it's not obvious that the optimal solutions are the same as they used to be.

[Author]

Posts